A Framework for Simulating Emergent Health Behaviors in Spatial Agent-Based Models of Disease Spread

Agent-based models (ABMs) offer valuable insights into disease dynamics and assist decision-makers in evaluating intervention strategies for disease outbreaks. Human health behaviors play a critical role in driving disease spread, and ABMs are well-suited to capture these behaviors. However, many existing ABMs of disease spread impose health behaviors onto agents and thus fail to consider the underlying drivers that give rise to health behaviors. Therefore, this study proposes a data-driven framework for modeling emergent health behaviors in ABMs of disease spread. The proposed framework is integrated into a geospatial ABM that simulates the spread of COVID-19 and mask use among the student population at George Mason University. The agent behavioral framework employs a logistic regression model to estimate the likelihood of an individual agent adopting a certain behavior at each time step. This estimation takes into account the individual's characteristics and perceived vulnerability to the disease, along with empirical data on the likelihood that such factors lead to behaviors like wearing a mask. Model results indicate a potential to not only use this approach to predict disease outcomes, but also to predict emergent patterns in health behaviors given the characteristics and perceptions of a given population. This modeling approach is flexible so that it can be used to simulate a range of health behaviors and diseases to provide better support for decision-makers in public health.